Bio-mimetic navigation in a dynamic environment

Publication Type:
Thesis
Issue Date:
2015
Full metadata record
Files in This Item:
Filename Description Size
Thumbnail01front.pdf94.21 kB
Adobe PDF
Thumbnail02whole.pdf1.63 MB
Adobe PDF
The importance of navigation in robotics cannot be understated. Without being able to correctly and efficiently navigate in an environment, an agent will be unable to carry out any tasks. Animals, even smaller mammals for instance, have sufficiently developed navigation systems that enable them to carry out their daily tasks: forage for food, find shelter, navigate to and fro such sites. It has long been proposed that animals use a type of map for navigation. Unlike maps generated by modern mapping techniques, these maps are topological, and lack precise metric information. Brain research has found sets of neurons that co-operate to form a navigation system in animals. Such cells: head direction cells, place cells, grid cells; decode specific information about the animal’s navigation, a combination of which is sufficient to provide a complete navigation solution. The aim of my masters research, as detailed in this report, was to study these spatial neurons and their modelling for use in robotic navigation. I have modelled head direction and grid cells, which are important components of the neural path integrator system using the Robot Operating System (ROS) platform. Both models have been validated with real time data collected from the PR2 robot.
Please use this identifier to cite or link to this item: